Reliability and controls
Billing Incident Runbooks: Protect Evidence Before Reprocessing
By the Orvlin team · · 1 min read
Billing incident response runbook: what to decide
A duplicated batch reaches invoices. The team’s first instinct is to delete it and rerun everything. That may fix one total while destroying the record of what customers saw and which documents require corrections.
Identify affected accounts, time range and pipeline stage. Pause risky downstream actions while retaining source records. Coordinate operational recovery with finance so meter repair and issued-document corrections remain linked.
An illustrative case
An incident duplicates usage for fifty accounts, but only ten invoices have been finalized. The recovery path differs for drafts and issued documents. Treating all fifty identically can create unnecessary credits or missing adjustments.
The important caveat
Do not communicate a final financial impact before reconciliation. Provide a factual initial notice, then update customers with confirmed scope and the remedy when the evidence is ready.
Review these decisions
- Contain invoice or credit side effects.
- Preserve original and processed records.
- Separate draft and finalized document handling.
- Reconcile and document preventive changes.
Frequently asked questions
- Should we rerun the entire billing period?
- Only when the scope and side effects are understood. A targeted, reviewed recovery can be safer than an unbounded replay.